Linksys Powerline A/V - Double check my assumptions

[weeber]

Hey all,

A little background, I've been running SageTV for over 2 years now on my ancient 1.4GHz Athlon and 27" SDTV. But I just bought my first house, and in addition to getting my first HDTV, I want to install a new Sage server/client setup that can stream HDTV.

The problem is, I have enough to do with this house that I don't want to run Cat5e cable throughout it and I don't trust wireless (even the high power pre-N, MIMO stuff). I decided to give the Linksys Powerline A/V product (networking through the existing powerlines in the house)a try, but first I wanted to test it. I found that this product was amazingly consistent in this house. I was able to play hi-def videos on my laptop from my HTPC, without hiccups, no matter where I plugged in. I also transferred a 454MB file and found it consistently took approximately 165 seconds to transfer [This was using Windows explorer to copy & paste the file, which I've heard is very inefficient]. By my calcs, this gives me a transfer rate of 22 Mbits/s. As a frame of reference, the same file takes 117 seconds to transfer when hard-wired with a 100Mbit/s Cat5 cable or 31 Mbits/s.

Based on the above, I'd just like to confirm that I have enough bandwidth for streaming HD video using the powerline product. It appeared to work fine and consistently, but I'm afraid I might be a little low. Just wanted to bounce this off some of the members here to see what they think. Should this be work for my needs? Or will I have to climb into my hot attic this summer to run Cat5 cable throughout the house?

Thanks

*Update 1*
Well, I just discovered that when I transfer the same file (via Cat5) between my main desktop and laptop it only takes 57 seconds, giving 64 Mbps. Whereas the transfer (also Cat5) between my HTPC and laptop is only 31 Mbps. Between this and the amazing consistent transfer rates I foundI'm beginning to think there's something wrong with the network setup on my HTPC. I have a feeling I can get more out of these things with a properly configured network setup. I'll play with it some more and let you know what I find.

[stevech]

are you in a No. American or other home with two-phase power (most all homes have that). I'd like to know if this new generation of ethernet on power line works across phase bridges. This means that the two network nodes are on opposite phases. The breaker box sets up which outlets are on which phase.

In X10 and other power line signaling, one usually has to put in a phase bridge coupler to provide a signal path from one phase to another. Another way to temporarily couple is to turn on any 220V appliance; it will bridge.

The symptom of a phase bridge problem is that certain pairs of wall outlets don't work reliably, or for the case at hand, operate at much reduced speeds due to weak signals.

I'm too cheap to buy these things at close to $200 per pair. My 11g WiFi does stream std def TV via Sage with no problems. I measured my plain old 11g at 24Mbps using the elapsed time it takes to transfer 2+GB hard disk image once a week at 2AM.

So if you have the new generation of power line ethernet, either it is no faster than 11g's speed, or your PC's disks or operating system is constraining the speed. You can try speed tests using freeware such as Ixia's Qcheck in TCP mode for max sized transfers.

[nielm]

22Mbits/sec = 9Gb/Hr

That should be just enough for SDTV, but maybe not enough for HDTV depending on your HDTV bitrate.

Bear in mind that some people have problems streaming SDTV over wireless 'g' (max 56Mbps) - it's not just the overall speed, but the consistancy. A 0.5 sec network dropout every minute or so means nothing to a file transfer, but will cause severe annoyance to video playback.

[weeber]

Stevech,

Yes, my house does have two phase power (that means 220V + 110V, right?).
The exact product that I have is the Linksys PLK200 starter kit.

Anyways, I tested this system by leaving my HTPC plugged into the living room and testing the laptop + other adapter in five different outlets/rooms throughout my house (2 downstairs and 3 upstairs). It was surprisingly consistent at 22Mbps throughout all the outlets, even when I had the dryer running for one test (I've heard electrical noise from a dryer can really impact these systems). It was so consistent that I'm thinking something else may have been the bottle-neck. My laptop is recent, but my HTPC is an old Athlon T-bird on a VIA KT133A chipset with a cheap PCI network adapter. However, things are slightly faster when I use a Cat5 cable, so maybe that's not it. I'll try that Qcheck program to see what I get next.

Yeah, I know it's not cheap but the only other alternatives as I see them are to spend just as much money on a whiz-bang MIMO, pre-N, high-power wireless router and multi-antennae reciever card. Or spend hours in the attic of my Atlanta, GA home (in the middle of summer) running Cat5 through the walls.

Nielm,

Yeah I was afraid that my measured bandwidth might be a hair low. I was actually hoping for a little better from this product, but it doesn't seem bad enough for me to return it yet. I'll play with it a little longer to see what I get.

[lobosrul]

22Mbits/sec should theoretically be enough for one HDTV stream. ATSC can never be over 19.3 Mbits/sec.

I never could get 1080i content to play over wireless (my 720p stations broadcast @ ~13mbps and did ok), so I simply run an ethernet cable down the floor next to the wall, but I'm not married. Low WAF I should imagine.

[weeber]

Yeah, WAF isn't a factor for me either, but I thought it'd be nice to try and step it up a bit and keep the house nice. Thus, no exposed cables.

I'm still debating what to do. Part of me wants to keep playing with these to see if I can get it any better, but another part says to just forget it and try something else. My results appear to be right on the border line of what's acceptable. If it were just a little bit slower or a little bit faster, the decision would be much easier.

[stevech]

Try speed at outlets that are likely on opposite phases when all your 220V appliances are OFF. Electric range/oven, electric dryer, some air conditioners, etc. When a 220V appliance is on, it bridges signals across the phases and benefits.

A big problem with power line carrier system are "signal suckers", like surge protector power strips and some electronics like PCs and new TVs. And especially UPSes. The surge suppressors in these attenuate the signal on the power line. The shorter the amount of wire between such a device and the wireless modem node, the more attenuation of the signal occurs.

Thanks for the initial reports. Not many folks are taking the time to report reality with these.

[mattdcknsn]

I am not using the linksys model but the Netgear version HDX101, currently I am getting around 90 Mbps Transmit and 85 Mbps receiving. I don't know if they are on the same side of the breaker box, since they worked easy from the get go I never had to check.

For me the powerline adapters were a lifesaver. I cant run the cables through the wall (in a townhouse) and the wife wont allow a cat5 cable be run up the stairs. So far I haven't had a major problem with them.

I have noticed that the age of the house can have a big influence on these things. I have read where older houses work better than newer houses.

[stevech]

what method did you use to measure the speeds?

[mattdcknsn]

the netgear app shows the speed it has between the different points. But I used a robocopy of a 2 gig file to see the actual usable speed. I had to rely on the netgear app to show me the return speed since I didn't want to do all of the tests twice.

[jdamore]

Originally Posted by mattdcknsn

I am not using the linksys model but the Netgear version HDX101, currently I am getting around 90 Mbps Transmit and 85 Mbps receiving. I don't know if they are on the same side of the breaker box, since they worked easy from the get go I never had to check.

For me the powerline adapters were a lifesaver. I cant run the cables through the wall (in a townhouse) and the wife wont allow a cat5 cable be run up the stairs. So far I haven't had a major problem with them.

I have noticed that the age of the house can have a big influence on these things. I have read where older houses work better than newer houses.

wow thats fast. I had mine in the same outlet and had <70 Mbps (using Qcheck). This could be that I am in an apartment with a very high noisefloor on the powerline. When going from living room to bed room I messured about 14 Mbps. This worked fine for me when playing sd video but, when the wife started to use the hair dryer it killed it.

[stevech]

Originally Posted by jdamore

his worked fine for me when playing sd video but, when the wife started to use the hair dryer it killed it.

Ah, the snag with power line data transmission.

[MattJackson86]

I have the linksys kit to allow me to play recordings (incuding HD) from my server to my pc upstairs. I found it does stutter at first (I guess due to lack of transfer speed) but i just pause the video for 10 secs or so (to allow it to create a buffer) and then it plays fine.